def plot_violin():
print "Plotting violin..."
plt.ioff()
fig = plt.figure()
sns.set(style="ticks")
d = [[1, 2, 3, 4, 2, 3, 2, 3], [3, 4, 2, 3, 5, 6, 7, 7, 7]]
rs = np.random.RandomState(0)
n, p = 40, 8
d = rs.normal(0, 1, (n, p))
d += np.log(np.arange(1, p + 1)) * -5 + 10
f, ax = plt.subplots()
sns.offset_spines()
sns.violinplot(d)
sns.despine(trim=True)
def plot_violin(): print "Plotting violin..." plt.ioff() fig=plt.figure() sns.set(style="ticks") d = [[1,2,3,4,2,3,2,3], [3,4,2,3,5,6,7,7,7]] rs = np.random.RandomState(0) n, p = 40, 8 d = rs.normal(0, 1, (n, p)) d += np.log(np.arange(1, p + 1)) * -5 + 10 f, ax = plt.subplots() sns.offset_spines() sns.violinplot(d) sns.despine(trim=True)
def graph_mult_data(file_list,xcolumn,ycolumn,plot_labels,x_label,y_label,title,file_title,colors=None,cfgs=None,seaborn=False):
if seaborn:
seaborn = test_seaborn()
import seaborn as sns
logger.info("unpacking data")
if colors==None:
datalist = [[numpy.loadtxt(filename),label] for filename,label in zip(file_list, plot_labels)]
else:
datalist = [[numpy.loadtxt(filename),label,color] for filename,label,color in zip(file_list, plot_labels,colors)]
if seaborn:
sns.set_style("ticks", rc={'font.family': 'Helvetica'})
sns.set_context("paper")
logger.info("plotting")
fig = plt.figure()
axis = fig.add_subplot(111)
if seaborn:
sns.offset_spines(fig=fig)
if colors==None:
for data,label in datalist:
axis.plot( data[:,xcolumn],data[:,ycolumn],label=label)
else:
for data,label,color in datalist:
axis.plot( data[:,xcolumn],data[:,ycolumn],label=label,color=color)
plot_markups(fig,axis,title,x_label,y_label,file_title,cfgs=cfgs,seaborn=seaborn)
def make_violin(self):
"""
Violin plots are made for the outliers over redshift. Each violin is a box plot, i.e.,
it depicts the probability density of the outliers for a given bin in redshift.
"""
from matplotlib.mlab import griddata
import matplotlib.pyplot as plt
import seaborn as sns
self.logger.info("Generating violin plot...")
ind = range(len(self.outliers))
rows = list(set(np.random.choice(ind,10000)))
self.logger.info("Using a smaller size for space ({0} objects)".format(self.reduce_size))
outliers = self.outliers[rows]
measured = self.measured[rows]
predicted = self.predicted[rows]
plt.figure()
bins = np.arange(0,self.measured.max()+0.1,0.1)
text_bins = ["{0}".format(i) for i in bins]
digitized = np.digitize(measured, bins)
outliers2 = (predicted - measured)/(measured+1)
violins = [outliers2[digitized == i] for i in range(1, len(bins))]
dbin = (bins[1]-bins[0])/2.
bins += dbin
final_violin, final_names = [], []
for i in range(len(violins)):
if len(violins[i]) > 1:
final_violin.append(violins[i])
final_names.append(bins[i])
pal = sns.blend_palette([self.color_palette, "lightblue"], 4)
sns.offset_spines()
ax = sns.violinplot(final_violin, names=final_names, color=pal)
sns.despine(trim=True)
ax.set_ylabel(r"$(z_{\rm phot}-z_{\rm spec})/(1+z_{\rm spec})$", fontsize=self.fontsize)
ax.set_xlabel(r"$z_{\rm spec}$", fontsize=self.fontsize)
ax.set_ylim([-0.5,0.5])
xtix = [i.get_text() for i in ax.get_xticklabels()]
new_xtix = [xtix[i] if (i % 2 == 0) else "" for i in range(len(xtix))]
ax.set_xticklabels(new_xtix)
for item in ([ax.xaxis.label, ax.yaxis.label]):
item.set_fontsize(self.fontsize)
for item in (ax.get_xticklabels() + ax.get_yticklabels()):
item.set_fontsize(self.fontsize-10)
ax.set_position([.15,.17,.75,.75])
self.kde_ax = ax
plt.savefig("PHOTZ_VIOLIN_{0}.pdf".format(self.family_name), format="pdf")
def make_violin(self):
"""
Violin plots are made for the outliers over redshift. Each violin is a box plot, i.e.,
it depicts the probability density of the outliers for a given bin in redshift.
"""
from matplotlib.mlab import griddata
import matplotlib.pyplot as plt
import seaborn as sns
self.logger.info("Generating violin plot...")
ind = range(len(self.outliers))
rows = list(set(np.random.choice(ind, 10000)))
self.logger.info("Using a smaller size for space ({0} objects)".format(
self.reduce_size))
outliers = self.outliers[rows]
measured = self.measured[rows]
predicted = self.predicted[rows]
plt.figure()
bins = np.arange(0, self.measured.max() + 0.1, 0.1)
text_bins = ["{0}".format(i) for i in bins]
digitized = np.digitize(measured, bins)
outliers2 = (predicted - measured) / (measured + 1)
violins = [outliers2[digitized == i] for i in range(1, len(bins))]
dbin = (bins[1] - bins[0]) / 2.
bins += dbin
final_violin, final_names = [], []
for i in range(len(violins)):
if len(violins[i]) > 1:
final_violin.append(violins[i])
final_names.append(bins[i])
pal = sns.blend_palette([self.color_palette, "lightblue"], 4)
sns.offset_spines()
ax = sns.violinplot(final_violin, names=final_names, color=pal)
sns.despine(trim=True)
ax.set_ylabel(r"$(z_{\rm phot}-z_{\rm spec})/(1+z_{\rm spec})$",
fontsize=self.fontsize)
ax.set_xlabel(r"$z_{\rm spec}$", fontsize=self.fontsize)
ax.set_ylim([-0.5, 0.5])
xtix = [i.get_text() for i in ax.get_xticklabels()]
new_xtix = [xtix[i] if (i % 2 == 0) else "" for i in range(len(xtix))]
ax.set_xticklabels(new_xtix)
for item in ([ax.xaxis.label, ax.yaxis.label]):
item.set_fontsize(self.fontsize)
for item in (ax.get_xticklabels() + ax.get_yticklabels()):
item.set_fontsize(self.fontsize - 10)
ax.set_position([.15, .17, .75, .75])
self.kde_ax = ax
plt.savefig("PHOTZ_VIOLIN_{0}.pdf".format(self.family_name),
format="pdf")
def main(options):
"""Main logic of the script"""
df = pd.read_table(options.input, header=None)
if options.clustered:
ndf = pd.DataFrame({'hybrids': Counter(df[3].tolist())})
top_mirnas = Counter(df[3].tolist())
else:
ndf = pd.DataFrame({'hybrids': Counter(df[6].tolist())})
top_mirnas = Counter(df[6].tolist())
ndf.hybrids = np.log(ndf.hybrids)
exp = {}
with open(options.expressions) as e:
for row in csv.reader(e, delimiter='\t'):
if row[0] != 'Mirnas/Samples':
exp[row[0].replace("(star)", "*")] = np.log(
np.mean([float(value) for value in row[1:]]))
ndf['expr'] = [exp[i] for i in ndf.index]
ndf = ndf[ndf['expr'] >= options.level]
# ndf = ndf[ndf['hybrids'] <= 7.0]
if options.verbose:
syserr("Number of miRNAs: %i\n" % len(ndf.index))
for val in top_mirnas.most_common()[:options.top]:
syserr("%s\t%i\n" % (val[0], val[1]))
sns.set(style='white', font='serif')
grid = sns.JointGrid('hybrids',
'expr',
data=ndf,
dropna=True,
size=8,
ratio=10,
space=0.2)
grid.plot_marginals(sns.distplot,
color="firebrick",
rug=False,
hist=True,
kde=False)
grid.plot_joint(sns.regplot,
scatter_kws={"color": "slategray"},
line_kws={
"linewidth": 1,
"color": "firebrick"
})
grid.annotate(stats.pearsonr,
stat="pearson r",
template="{stat} = {val:.2g}\np-value = {p:.2g}")
grid.set_axis_labels('log(Number of hybrids)', 'log(Expression value)')
pl.savefig("hybrids_vs_expression.png")
pl.clf()
sns.set(style='ticks', font='serif')
sns.distplot(ndf.hybrids, kde=False)
sns.despine(trim=True)
sns.offset_spines()
pl.xlabel("log(Number of hybrids)")
pl.ylabel("Fraction")
pl.title("Number of hybrids found per miRNA")
pl.tight_layout()
pl.savefig("hist_of_hybrid_number.png")
pl.clf()
sns.set(style='ticks', font='serif')
sns.distplot(ndf.expr, kde=False)
sns.despine(trim=True)
sns.offset_spines()
pl.xlabel("log(Expression value)")
pl.ylabel("Fraction")
pl.title("Expression of miRNAs")
pl.tight_layout()
pl.savefig("hist_of_expressions.png")
""" Violinplots =========== """ import numpy as np import seaborn as sns import matplotlib.pyplot as plt sns.set(style="ticks") rs = np.random.RandomState(0) n, p = 40, 8 d = rs.normal(0, 1, (n, p)) d += np.log(np.arange(1, p + 1)) * -5 + 10 f, ax = plt.subplots() sns.offset_spines() sns.violinplot(d) sns.despine(trim=True)
